Hybrid Grey-Wolf Optimizer and MLP Neural Network for Shear Capacity Prediction of Slender FRP-RC Beams

N Zafarani and Y Sharifi, INTERNATIONAL JOURNAL OF CIVIL ENGINEERING, 23, 2341-2359 (2025).

DOI: 10.1007/s40999-025-01155-4

This research introduces an innovative hybrid model that integrates the Grey-Wolf Optimizer (GWO) with a Multilayer Perceptron (MLP) Neural Network (GWO-ANN) to predict the shear strength of slender fiber- reinforced polymer reinforced concrete (FRP-RC) beams without stirrups. Leveraging a comprehensive experimental dataset, the model produces a simple yet highly accurate closed-form equation for evaluating shear capacity. Various existing equations from codes and previous studies were reviewed and compared using carefully selected performance metrics. The comparative assessment demonstrates that the proposed GWO-ANN model outperforms current approaches by delivering improved precision without added complexity. This technique offers a valuable and reliable tool for engineers during the design and planning phases of structural projects. Notably, among existing design provisions, the Canadian design guidelines for fiber-reinforced polymer (FRP) reinforcement in concrete structures (ISIS M03-07) was identified as providing relatively more accurate predictions compared to other code equations.

Return to Publications page